LynxAI
Model guide
A plain-English guide to the two types of AI model available in LynxAI — and how to pick the right one for your task.
Standard Models
Fast and capable — the right choice most of the time
Standard models predict the next best response based on what they've learned. They're highly capable and respond quickly, making them the right default for the vast majority of tasks.
Best for
- Writing and editing — emails, briefs, reports, copy
- Summarising documents and meeting notes
- Translating content
- Brainstorming and ideation
- Answering questions and everyday chat
In LynxAI
- Claude Sonnet (default)
- Claude Haiku (fast, lightweight)
- Gemini Flash
Trade-off
Occasionally less reliable on very complex multi-step reasoning or advanced maths.
Reasoning Models
Slower and more deliberate — for your hardest problems
Reasoning models work through problems step-by-step internally — checking their own logic and correcting mistakes before showing you an answer. Think of them as the model that pauses to think before it speaks.
Best for
- Complex logic and multi-step analysis
- Strategic planning and scenario evaluation
- Reviewing assumptions and stress-testing arguments
- Advanced coding and debugging
In LynxAI
- Claude Sonnet with Extended Thinking enabled
- Gemini Pro
Trade-off
Slower response times and higher token cost — don't use for simple tasks.
The most powerful model isn't always the right one
Matching the model to the task is a skill — and an important one. A lightweight model handles most everyday work just as well as a frontier one, but at a fraction of the cost and in less time. When the whole team picks thoughtfully, LynxAI stays fast, available, and affordable for everyone. Save the heavy-duty reasoning models for the problems that genuinely need them.